sfit_minimizer.mm_funcs module

sfit_minimizer.mm_funcs.fit_mulens_event(event, parameters_to_fit=None, initial_guess=None, plot=False, tol=1e-05, verbose=False)

Basic function for fitting a point-lens microlensing model to data.

Arguments:
event: MulensModel.Event() object

event contains datasets and an initial model.

parameters_to_fit: list of str, optional

List of names of point-lens parameters to fit, e.g. [‘t_0’, ‘u_0’, ‘t_E’, ‘rho’]. If not given, will fit for all parameters of event.model.

initial_guess: list, np.ndarray, optional

Starting values for the parameters to be fitted. Must include values for the flux parameters. If not included, will be set to the values of the parameters in event.model and the initial values of the fluxes will be set with event.fit_fluxes().

plot: bool, optional

After the fitting is complete, plot the best-fitting model with the data.

tol: float

See sfit_minimizer.sfit_minimize.minimize().

verbose: bool

See sfit_minimizer.sfit_minimize.minimize().

Returns:

sfit_minimizer.sfit_classes.SFitResults object.

class sfit_minimizer.mm_funcs.PointLensSFitFunction(event, parameters_to_fit, estimate_fluxes=False, add_2450000=False)

Bases: SFitFunction

A class for fitting a point-source point-lens microlensing light curve to observed data using sfit_minimizer.sfit_minimize.minimize(). Simultaneously fits microlensing parameters and source and blend fluxes for each dataset.

Arguments:
event: MulensModel.Event() object

event contains datasets and an initial model.

parameters_to_fit: list of str

list of the named model parameters to be fit. (Not including the fluxes.)

estimate_fluxes: bool

If set, will use event.fit_fluxes() to generate initial values for the flux parameters for each data set. Otherwise, will initialize the fitting with fixed values from event.fix_source_flux and event.fix_blend_flux or if they are not set for a particular dataset, will use source_flux = 1. and blend_flux = 0.

add_2450000: bool

see add_2450000

Attributes:
n_params = int

Number of parameters to fit. Includes flux parameters.

Notes:

  1. if you want to fix the source or blend flux for a particular dataset,

use the fix_source_flux or fix_blend_flux keywords in event as usual.

  1. If u_0 is a parameter of the fit and it is too close to zero,

the matrix inversion will fail (produce numpy.linalg.LinAlgError: Singular matrix). So if you see this error, check the value of u_0. (Probably true of other parameters as well)

property add_2450000

bool

if True: add 2450000 to theta[t_0] before putting it in the model.

flatten_data()

Concatenates good points for all datasets into a single array with columns: Date, flux, err.

set_flux_indices()

Count the total number of parameters of the model, n_params, and sets makes list of indices to match the flux parameters to the correct columns in the b, c, and d matrices after accounting for fixed flux parameters.

update_all(theta=None, verbose=False)

See sfit_minimizer.sfit_classes.SFitFunction.update_all().

calc_residuals()

Calculate expected values of the residuals

calc_df()

Calculate the derivatives of the fitting function and store as self.df.